Savorant Research

Open studies.

Most of what people believe about tasting whiskey has never been tested on the people doing the tasting. Savorant records blind results, words and ratings from real tasters over months, which makes questions answerable that panels in a lab can’t reach. These are the questions we are asking, with the full plan for each published before any data is analyzed.

One taster, many nightsOpen

Is the triangle test a dependable measure for one person scoring themselves?

Target: 97 taster-pairs with at least 36 rounds

Water in the glassOpens with the Water Course

Does the water you add change what tasters notice in a whiskey?

Target: At least 109 water triangles from at least 25 tasters

Does training show up blind?Open

As a taster’s vocabulary grows, do their blind results change?

Target: 34 tasters with at least 54 rounds

Same whiskey, same words?Open

Is describing the same whiskey the same way twice a stable, measurable skill?

Target: 36 tasters with at least six repeat checks

Order and fatigueOpen

Does the order of the glasses change what tasters report?

Target: At least 252 oak-grip ratings at position 1 and 252 at position 3 or later

Perception over timeOpen

Does a taster’s perception stay the same while their vocabulary grows?

Target: 20 tasters with two scores at least 11 months apart

Glass shuffle or pourer?Open

Does blinding yourself with the glass shuffle give the same results as having someone pour?

Target: 35 tasters with at least 18 rounds in each mode on the same pair

Time of dayOpen

Does the time of day you taste change your blind results, and does tasting at the same time make them steadier?

Target: At least 392 daytime and 392 evening triangle rounds

How a study works

Two consents, both yours to give. Research contribution, in the app’s Settings → Research, lets your sessions be used for research at all, under a coded ID. Each study then asks on its own, with its own text: what it’s for, what you do, what it uses, the risks, and who runs it. You can join one, some or none, and leave any of them with one tap.

Earlier data asks separately. Joining covers what you file from then on. Where a study could use what you filed before you joined, it asks that as its own question, and the main result never depends on it.

Every result is published, as group results under coded IDs, including the ones that show nothing.

Standards every study follows

Preregistration. This page is the protocol of record, versioned and dated, and is posted to the Open Science Framework (OSF) before any study data is analyzed. Hypotheses, outcomes, sample size and analysis are fixed here in advance.
Primary and secondary outcomes. Each study names one primary outcome, tested at the stated alpha. Everything else is labeled secondary or exploratory and reported as such, with effect sizes and 95% confidence intervals rather than bare p-values.
No peeking. No interim analysis of the outcomes. The only checks before the sample size is reached are for data quality (duplicates, impossible values), and they never look at results.
Repeated measures. One taster contributes many rounds, so analyses use mixed-effects models with the taster as a random effect, or per-taster summaries, never pooled rounds treated as independent.
Exact methods for small counts. Triangle results are analyzed against the exact binomial with chance at one in three, the way ISO 4120 sets out the triangle test, adapted here for repeated measures.
Exclusions. Rounds are excluded only for the reasons listed in advance: test accounts, the founder’s accounts, rounds filed in a preview build, and duplicates. Every exclusion is counted and reported.
Missing data. Tasters who stop early are kept for every analysis their data supports. The number who stop, and when, is reported.
Deviations. Any change after data collection starts is logged on this page as a dated amendment with its reason, and reported in the paper.
Reporting. Results are reported following the STROBE guideline for observational studies, with the analysis code published alongside.
Conflict of interest. The study is run by the company that sells the app. Preregistration, publishing regardless of outcome, and public analysis code are how that interest is kept out of the result.

Where we are honest about where we are. No study has yet been reviewed by an independent research ethics board. We intend to have each one reviewed before any results are published, and if that review changes what a study asks of you, the app will ask you again. The studies are run by John Schuster for Drunken Unicorn Productions, the company that makes and sells Savorant.

Questions, or a method you think is wrong: palate@savorant.com. We will answer, and if we change a protocol, the change is logged on its page with the date and the reason. The evidence review covers what is already published.